:format(webp)/cdn.vox-cdn.com/assets/1115546/bball_dictionary.jpg)
The crickets are getting restless. The draft lottery is still a few weeks away, and while many Warriors fans have probably taken the summer off already, hardcore GSoMers like you are in need of some nourishment to keep you going. I'm here to satisfy a little bit of that hunger with some advanced stats, and not your run-of-the-mill TS% or USG%. I trust that if you're reading this now, you're smart and resourceful enough to look up those stats on your own. Heck, I'm sure you already have. No, even so-called "advanced stats" are not good enough for GSoM. In this post, I'm going to give a rundown of Synergy and my own metric Adjusted Four Factor +/- (A4PM), which breaks down RAPM into a player's effects on team shooting efficiency, turnover rates, rebounding, and free throw attempts.
Synergy
Recall that the video service Synergy breaks down plays into these categories (my abbreviations in parentheses):
- Isolation (ISO)
- Post-up (POST)
- Pick and roll - ball handler shoots (BALL)
- Pick and roll - rolling man shoots (MAN)
- Spot-up jump shot (SPOT)
- Coming off screen (SCREEN)
- Hand off to shooter (HAND)
- Cutting to basket (CUT)
- Following offensive rebound (OREB)
- Shot coming in transition (TRANS)
- Other not categorized (OTHER)
Note that in these tables I only show players that qualified with enough plays to have a Synergy ranking. That's why, for example, you won't find Andris Biedrins anywhere.
ISO
Overall, the Warriors ran ISO 12.8% of the time with a 0.76 PPP (ranked #19 in the league).
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Brandon Rush |
ISO |
10.60% |
60 |
0.87 |
42 |
|
David Lee |
ISO |
11.80% |
139 |
0.86 |
46 |
|
Nate Robinson |
ISO |
18.00% |
108 |
0.84 |
60 |
|
Klay Thompson |
ISO |
10.30% |
88 |
0.77 |
104 |
|
Stephen Curry |
ISO |
15.10% |
56 |
0.75 |
123 |
|
Monta Ellis |
ISO |
22.60% |
201 |
0.73 |
142 |
|
Charles Jenkins |
ISO |
14.00% |
50 |
0.72 |
148 |
|
Richard Jefferson |
ISO |
13.40% |
27 |
0.67 |
177 |
|
Dorell Wright |
ISO |
7.60% |
47 |
0.60 |
201 |
|
Dominic McGuire |
ISO |
15.50% |
46 |
0.57 |
208 |
Isolation is a play that teams should generally try to avoid because is is not that efficient, although ironically, many of the league's star players are recognized for their ability in isolation (Melo, Kobe, etc.). Without having Monta Ellis around as an "iso crutch", the Warriors will most likely run this play less often - but that could be a blessing in disguise. And when MJ does feel the need to call ISO, the Warriors actually appear to have several decent options. It's nice to see Brandon Rush at the top of this list. He's known mostly for his spot-up shooting, but apparently, when he puts the ball on the floor, he is able to create his own shot and score efficiently. It's a relatively small sample size, though, and the previous season his efficiency on ISO was only 0.70. David Lee, on the other hand, was even more efficient in isolation last season (0.90 PPP), so that number looks believable. Stephen Curry was actually much better in ISO last season (0.93 PPP on 197 plays). He only ran 56 ISO plays this season, which is a relatively small sample, but my guess is that his ankle problems had something to do with his poorer performance this season.
SPOT
The Warriors ran SPOT 19.1% of the time for a 1.09 PPP which ranked #1 in the league. This is clearly our wheelhouse.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Stephen Curry |
SPOT |
21.40% |
79 |
1.38 |
3 |
|
Richard Jefferson |
SPOT |
37.80% |
76 |
1.37 |
5 |
|
Monta Ellis |
SPOT |
9.80% |
87 |
1.29 |
9 |
|
Brandon Rush |
SPOT |
34.90% |
198 |
1.22 |
17 |
|
Klay Thompson |
SPOT |
19.70% |
168 |
1.13 |
35 |
|
Nate Robinson |
SPOT |
18.10% |
109 |
1.11 |
41 |
|
Dorell Wright |
SPOT |
40.70% |
251 |
1.07 |
62 |
|
Charles Jenkins |
SPOT |
22.10% |
79 |
0.95 |
155 |
|
David Lee |
SPOT |
12.10% |
143 |
0.93 |
171 |
|
Dominic McGuire |
SPOT |
11.50% |
34 |
0.56 |
331 |
I know what you're thinking...Monta shot that efficiently? The year before his PPP on SPOT plays was only 1.01. In his short time with the Bucks, he shot 1.0 PPP in 50 attempts. This just shows you (like Brandon Rush's ISO efficiency above) to be careful when looking at these Synergy numbers. These numbers bounce around, sometimes quite a lot. It's pretty clear, though, that the Warriors have assembled a team with a bunch of good spot-up shooters. The addition of Bogut could even help space the floor more and make our perimeter shooting even better. That's the hope, anyway. Before moving on, I just have to point out one thing. Dominic McGuire. Dude really can't shoot. He may be the only player in the NBA who shot better in isolation than as a spot-up shooter, which is hilarious considering how bad his ISO efficiency was (0.57 PPP). Just saying. Dude who can't shoot.
POST
The Warriors ran this play 9.4% of the time for 0.84 PPP which ranked #15 in the league.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Monta Ellis |
POST |
7.70% |
69 |
1.07 |
6 |
|
Ekpe Udoh |
POST |
37.50% |
93 |
0.84 |
68 |
|
David Lee |
POST |
23.70% |
280 |
0.83 |
73 |
|
Jeremy Tyler |
POST |
30.40% |
80 |
0.83 |
73 |
|
Klay Thompson |
POST |
3.10% |
26 |
0.77 |
105 |
How funny is it that the Warriors, generally known as a team with a woeful post offense, lose their two best post players, one of who is a small combo guard and the other has been referred to as a "No-Stats All-Star"? Yup, that's our Warriors. The hope is Andrew Bogut helps us here, but it should be noted that his POST efficiency this season was only 0.71 PPP. In 2010-11, his efficiency was somewhat better at 0.78 PPP. And going back to 2009-10, his efficiency was 0.80 PPP. Bogut certainly demands attention in the post, but he's not what you would call a high efficiency option down low. In Milwaukee he was asked to generate a lot of offense, and one would think that with the shooters we have, Bogut won't be counted on to generate nearly as much offense. But I think it's fair to say he doesn't solve all our problems down low. If you're expecting Dwight Howard type post offense (0.88 PPP on 57.4% of his plays), you may be sorely disappointed.
BALL
The Warriors ran this play 13.1% of the time for 0.79 PPP which ranked #17.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Brandon Rush |
BALL |
5.60% |
32 |
0.97 |
14 |
|
Stephen Curry |
BALL |
25.70% |
95 |
0.97 |
14 |
|
Charles Jenkins |
BALL |
35.30% |
126 |
0.93 |
25 |
|
Nate Robinson |
BALL |
35.90% |
216 |
0.82 |
58 |
|
Monta Ellis |
BALL |
28.30% |
252 |
0.73 |
103 |
|
Dorell Wright |
BALL |
4.90% |
30 |
0.70 |
112 |
|
Klay Thompson |
BALL |
9.80% |
83 |
0.52 |
177 |
Last season Curry shot 0.89 PPP on the pick and roll play, and 0.81 PPP his rookie season. It appears he has steadily improved here, which is a great sign. The fact that Jenkins is already at 0.93 PPP (on 35%) is also very encouraging. If he can just add a 3-pt shot to his arsenal, he would be a much more efficient scorer, overall. In fact, CJ was quite a bit more efficient in the pick and roll than Kyrie Irving (0.82 PPP). It should be noted that the best PNR players run this play a lot more. 61% of Nash's shots came on this play (0.93 PPP). Chris Paul ran this play 41% of the time (0.94 PPP). As Curry and/or Jenkins run this play more, they will (deservedly) receive more defensive attention, so their efficiencies may decrease somewhat.
MAN
The Warriors ran this play 5.6% of the time for a 1.04 PPP which ranked #5.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Ekpe Udoh |
MAN |
10.50% |
26 |
1.23 |
10 |
|
David Lee |
MAN |
19.30% |
228 |
1.03 |
48 |
|
Jeremy Tyler |
MAN |
10.30% |
27 |
0.74 |
125 |
This is where David Lee really helps. Pick your poison on the pick and roll, either Curry (Jenkins) or Lee. When the Warriors traded for Lee, one of the thoughts was that the PNR would become a bread-and-butter play for us. Now that Curry will be running the show in the backcourt, the odds of that happening are much higher. Regarding Udoh's apparently high efficiency, the sample size should be taken into account. In his short time with the Bucks, he shot 0.92 PPP on 37 pick and roll plays. That's not bad, though. Finally, it should be obvious that Tyler needs to improve in this area.
SCREEN
The Warriors ran this play 7.2% of the time at 0.92 PPP which ranked #10.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Stephen Curry |
SCREEN |
10.00% |
37 |
1.05 |
14 |
|
Klay Thompson |
SCREEN |
28.10% |
239 |
1.03 |
16 |
|
Monta Ellis |
SCREEN |
6.60% |
59 |
0.83 |
75 |
|
Brandon Rush |
SCREEN |
8.80% |
50 |
0.80 |
83 |
|
Dorell Wright |
SCREEN |
8.30% |
51 |
0.59 |
119 |
It's pretty easy to predict the Warriors offensive scheme next season. Pick and roll with Curry and Lee. Klay Thompson coming off screens set by Lee or Bogut. Brandon Rush parking somewhere on the 3-pt line and keeping defenses honest. And on occasion, running a post play for Bogut.
TRANSITION
The Warriors ran this play 13.2% of the time at 1.15 PPP which ranked #9.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Dorell Wright |
TRANS |
20.90% |
129 |
1.26 |
77 |
|
Brandon Rush |
TRANS |
21.00% |
119 |
1.26 |
77 |
|
Nate Robinson |
TRANS |
15.10% |
91 |
1.25 |
84 |
|
David Lee |
TRANS |
6.30% |
75 |
1.24 |
89 |
|
Monta Ellis |
TRANS |
11.90% |
106 |
1.22 |
103 |
|
Klay Thompson |
TRANS |
15.30% |
130 |
1.12 |
165 |
|
Stephen Curry |
TRANS |
17.00% |
63 |
1.05 |
206 |
|
Dominic McGuire |
TRANS |
16.60% |
49 |
1.00 |
226 |
|
Charles Jenkins |
TRANS |
13.70% |
49 |
0.73 |
284 |
Yay, Dorell wins a category! Ok, the only thing to note here is that Curry, Klay, and especially Jenkins need to pass more on transition plays, amirite?
CUT
The Warriors ran this play 7.2% of the time at 1.21 PPP which ranked #12.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Brandon Rush |
CUT |
4.90% |
28 |
1.46 |
12 |
|
David Lee |
CUT |
11.30% |
133 |
1.38 |
21 |
|
Monta Ellis |
CUT |
4.80% |
43 |
1.30 |
46 |
|
Dorell Wright |
CUT |
5.70% |
35 |
1.20 |
95 |
|
Dominic McGuire |
CUT |
19.30% |
57 |
1.11 |
140 |
|
Ekpe Udoh |
CUT |
13.30% |
33 |
0.97 |
186 |
|
Jeremy Tyler |
CUT |
16.00% |
42 |
0.79 |
207 |
The Warriors have several guys who can finish around the basket. Right now, Jeremy Tyler doesn't appear to be one of those guys.
OREB
5.1% of the Warriors' plays came off of offensive rebounds at a 1.01 PPP which ranked #23.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Dorell Wright |
OREB |
5.20% |
32 |
1.34 |
3 |
|
David Lee |
OREB |
8.60% |
102 |
1.13 |
55 |
|
Jeremy Tyler |
OREB |
15.60% |
41 |
0.95 |
123 |
|
Ekpe Udoh |
OREB |
10.90% |
27 |
0.89 |
140 |
|
Dominic McGuire |
OREB |
12.20% |
36 |
0.47 |
158 |
This is another area where Bogut should help, and yet another area where Tyler could use some improvement. And seriously, Dominic McGuire. #smh
HAND
This is a rarely used play (i.e. usually as the shot clock winds down). The Warriors ran this 1.7% of the time at 0.85 PPP which ranked #12.
|
NAME |
PLAY |
%TIME |
NUM |
PPP |
RANK |
|
Klay Thompson |
HAND |
3.80% |
32 |
0.81 |
35 |
A4PM
For a complete description of A4PM see here and here. Briefly, the main idea is that I calculate adjusted versions of the four factors (eFG%, TOR, ORB%, FTA/FGA) on offense and defense, and then add up all those individual components to create a composite version of RAPM. One advantage of this technique is that, as opposed to RAPM, it is not as much of a "black box", since you can see the player's effect on the four factors. A second advantage is that it appears to be more stable year-to-year (as shown here), which suggests it's predictive capability may be better than RAPM (although I haven't proven that yet).
Here, I'll look at the 4 offensive components and the total offensive rating (O4PM).
O4PM
|
NAME |
POSS |
O4PM |
|
Nate Robinson |
2212 |
1.66 |
|
Stephen Curry |
1417 |
1.23 |
|
Klay Thompson |
2822 |
0.75 |
|
Brandon Rush |
3149 |
0.37 |
|
David Lee |
3915 |
0.26 |
|
Dorell Wright |
3047 |
-0.03 |
|
Richard Jefferson |
3072 |
-0.21 |
|
Dominic McGuire |
1976 |
-0.48 |
|
Jeremy Tyler |
878 |
-0.57 |
|
Andris Biedrins |
1345 |
-0.66 |
|
Charles Jenkins |
1487 |
-1.79 |
For the curious, Monta's O4PM was 0.22 this season, and Udoh's was 0.51.
eFG%
The OEFG column lists the adjusted eFG% value (in %-points) and RATING represents the contribution of that value to O4PM (in this case 1.77*eFG%).
|
NAME |
POSS |
RATING |
OEFG |
|
Stephen Curry |
1417 |
1.45 |
0.82 |
|
Nate Robinson |
2212 |
1.08 |
0.61 |
|
Dominic McGuire |
1976 |
0.95 |
0.53 |
|
Brandon Rush |
3149 |
0.87 |
0.49 |
|
Klay Thompson |
2822 |
0.68 |
0.38 |
|
David Lee |
3915 |
0.32 |
0.18 |
|
Andris Biedrins |
1345 |
0.26 |
0.15 |
|
Dorell Wright |
3047 |
0.13 |
0.07 |
|
Richard Jefferson |
3072 |
-0.24 |
-0.14 |
|
Jeremy Tyler |
878 |
-0.31 |
-0.18 |
|
Charles Jenkins |
1487 |
-1.33 |
-0.75 |
OTOR
This is the adjusted turnover rate per 100 possessions (negative values are better, meaning fewer turnovers). The contribution to O4PM is -1.37*OTOR.
|
NAME |
POSS |
RATING |
OTOR |
|
Nate Robinson |
2212 |
0.79 |
-0.58 |
|
Richard Jefferson |
3072 |
0.69 |
-0.50 |
|
Klay Thompson |
2822 |
0.65 |
-0.48 |
|
David Lee |
3915 |
0.31 |
-0.22 |
|
Stephen Curry |
1417 |
0.27 |
-0.20 |
|
Dorell Wright |
3047 |
0.11 |
-0.08 |
|
Brandon Rush |
3149 |
0.10 |
-0.07 |
|
Andris Biedrins |
1345 |
0.06 |
-0.05 |
|
Charles Jenkins |
1487 |
-0.09 |
0.07 |
|
Jeremy Tyler |
878 |
-0.28 |
0.21 |
|
Dominic McGuire |
1976 |
-0.78 |
0.57 |
ORR
The value given for ORR is the adjusted %-change in offensive rebounding rate. The contribution to O4PM is 0.31*ORR. No surprise here, we are not a good rebounding team.
|
NAME |
POSS |
RATING |
ORR |
|
Jeremy Tyler |
878 |
0.11 |
0.35 |
|
David Lee |
3915 |
-0.15 |
-0.49 |
|
Stephen Curry |
1417 |
-0.17 |
-0.55 |
|
Dorell Wright |
3047 |
-0.20 |
-0.65 |
|
Charles Jenkins |
1487 |
-0.22 |
-0.71 |
|
Nate Robinson |
2212 |
-0.31 |
-1.01 |
|
Klay Thompson |
2822 |
-0.36 |
-1.17 |
|
Brandon Rush |
3149 |
-0.42 |
-1.34 |
|
Dominic McGuire |
1976 |
-0.51 |
-1.64 |
|
Richard Jefferson |
3072 |
-0.55 |
-1.79 |
|
Andris Biedrins |
1345 |
-0.58 |
-1.87 |
FTA/FGA
The value of OFTR is in terms of FTA per 100 FGA. The contribution to O4PM is 0.14*OFTR.
|
NAME |
POSS |
RATING |
OFTR |
|
Nate Robinson |
2212 |
0.10 |
0.74 |
|
Dorell Wright |
3047 |
-0.08 |
-0.54 |
|
Jeremy Tyler |
878 |
-0.08 |
-0.61 |
|
Richard Jefferson |
3072 |
-0.10 |
-0.73 |
|
Dominic McGuire |
1976 |
-0.14 |
-0.97 |
|
Charles Jenkins |
1487 |
-0.15 |
-1.08 |
|
Brandon Rush |
3149 |
-0.19 |
-1.35 |
|
David Lee |
3915 |
-0.21 |
-1.49 |
|
Klay Thompson |
2822 |
-0.22 |
-1.57 |
|
Stephen Curry |
1417 |
-0.31 |
-2.24 |
|
Andris Biedrins |
1345 |
-0.41 |
-2.90 |
I'll be happy to answer any questions you have in the comments section. In Part 2, I'll go over the defense.